A Power-aware Task Scheduling Algorithm for Application of DVS on Multi-processor Systems

DOI Web Site Web Site 21 References
  • Mori Yuichiro
    Department of Systems and Social Informatics, Graduate School of Information Science, Nagoya University
  • Asakura Koichi
    Department of Information Systems, School of Informatics, Daido University
  • Watanabe Toyohide
    Department of Systems and Social Informatics, Graduate School of Information Science, Nagoya University

Bibliographic Information

Other Title
  • マルチ・プロセッサ環境におけるＤＶＳを用いた消費電力量削減アルゴリズムの構築
  • マルチ プロセッサ カンキョウ ニ オケル DVS オ モチイタ ショウヒ デンリョクリョウ サクゲン アルゴリズム ノ コウチク

Search this article

Abstract

In this paper, we propose a power-aware scheduling algorithm for reducing power consumption especially for highly-loaded processors without increasing the makespan of programs. In our algorithm, Dynamic Voltage Scaling (DVS) is applied to tasks which are not on the critical path of a schedule. In order to find tasks which can be applied DVS, we introduce a concept of slack-time. Slack-time represents the amount of redundant time for delaying completion time of a task without increasing the makespan of the schedule. In our algorithm, firstly, slack-time is calculated for each task. Then, tasks are selected based on a greedy method. In this mechanism, DVS is applied to tasks that have slack-time in decreasing order of reducing power consumption. Experimental results show that our algorithm can reduce about 10.6% and 16.5% of power consumption on 4 and 8 processors respectively on average.

Journal

References(21)*help

See more

Keywords

Details 詳細情報について

Report a problem

Back to top